Psychological Stress Linked with Autoimmune Disease

A Swedish study of more than 1 million individuals suggests that stress disorders like PTSD increase the risk of autoimmune disease later in life.

 Sweden. I'm like, 99% sure.

Sweden. I'm like, 99% sure.

Ah Sweden, land of the midnight sun, delicious cherry-flavored fish, and a national medical registry that captures every single living person in the country.

Leveraging Sweden’s robust patient registry is something of a right of passage for epidemiologists looking for population-level evidence of disease association, and this paper, appearing in the Journal of the American Medical Association, found one that seems to be fairly compelling.

The study asked the question of whether psychiatric stress-related disorders like post-traumatic stress disorder are linked to autoimmune diseases (like Lupus and Hashimoto’s thyroiditis – actually a total of 41 autoimmune diseases. That’s the power of big data).

Overall, the authors identified 106,464 Swedes who had suffered from a stress-related disorder and age- and sex-matched them to 1,064,640 Swedes who had not. They then looked to see, over a period of almost 30 years, how many would develop autoimmune diseases.

The results look like this:

 Rates of autoimmune diseases stratified by history of prior stress disorder.

Rates of autoimmune diseases stratified by history of prior stress disorder.

There’s a slightly higher rate of autoimmune disease in those who had a prior stress disorder. Take my word for it, it’s a slightly bigger piece of pie. How much bigger? Well, if you took 1000 people with stress disorders, 9 would develop an autoimmune condition over the next year, compared to 6 out of 1000 of those without stress disorders.

This modest increased risk persisted even after accounting for differences in marital status, comorbidities, history of psychiatric disorders, and educational level.

Why?

 Effect mediated by "cytokines"

Effect mediated by "cytokines"

Well, the authors argue that these stressful episodes might lead to alterations in the immune system via a variety of mechanisms (none were directly studied in this study) such as cortisol depression or upregulation of inflammatory cytokines.

 Effect mediated by alcohol

Effect mediated by alcohol

You and I might argue that unmeasured factors like alcohol intake and smoking might be equally important here, either as mediators of the observed link

 Confounding by smoking

Confounding by smoking

(maybe people who experience a significant stress start drinking more which may lead to autoimmunity) or as a confounder

 (maybe people who smoke are more likely to have stress reactions and more likely to have autoimmunity and the two factors aren’t linked at all).

The authors note that given the small absolute risk of autoimmunity, we shouldn’t be screening patients with stress-disorders for autoimmune disease. But the study did observe at least one interesting effect that needs further exploration. Those patients who used SSRIs for stress disorders in this study had a substantially higher rate of autoimmune disease, but the LONGER they used SSRIs, the lower the risk became.

I asked lead author Dr. Huan Song about this paradoxical finding. She told me she suspects that SSRIs are a marker for severity of the stress-disorder, leading to the higher effect size, but that the persistent use of SSRIs could be beneficial.  Perhaps they modify some of the effects of stress. This means there’s a potential therapeutic target here. When will we know for sure whether this link represents real physiology and not statistical artifact? Don’t stress – those studies are ongoing.